Food demand is expected to be 350 million tonnes by 2030 and we are currently producing 291.95 million tonnes ( 2019-20 ).
Average crop productivity in our country is low due to lack of infrastructure and proper supply chains.
In Agriculture, there is a great potential for AI and ML systems to improve the efficiency and quality of food crops
It is based on following parameters:
● Nitrogen
● Phosphorous
● State
● City
● pH
● Rainfall
State and City parameters are used by the Weather API to get the current temperature, rainfall, Humidity details for the specific location. All these parameters are fed to the Model for prediction Here, the Random Forest algorithm gives us the maximum accuracy. Based on prediction, Output is displayed on screen
Existing Fertilizer dataset is used and this is used as a training and testing datasets. User inputs
● Nitrogen
● Phosphorous
● Soil Type
Weather Details ( Temperature, Humidity, Moisture ) are fetched via API call.
Based on data given, output box is displayed on screen having details such as
● Conventional recommended fertilizer
● Organic alternatives to this fertilizer
● General Information
● Dosage for crops
Each year, the Government releases a price index called WPI (Wholesale price Index). Wholesale price indexes are reported monthly in order to show the average price changes of goods.
To get the current price for the current year, we use the following formula Current price = WPI * Base Price(for the ongoing year)
Using previous year data, predictions have been made to indicate what might be the
● maximum price
● average predicted price
● minimum price
Graph is also plotted. Left graph shows the projected price while the right graph shows the history of the crop prices from 2012 to 2019.
Sarimax Time series forecasting model is being used for the prediction of the future crop prices. This model can predict prices upto 4-5 years accurately.
The farmer has to upload an image to the website from his/her device. Deep Learning models are used is to find the disease that occurred and details about the disease and its cure are given to the farmer.
Farmers can stay up to date with the latest news in the agricultural domain with help of the news portal. The news is dynamically fetched and is updated every day.
From various sites, news is gathered via web crawling and then displayed on the news portal.
On Clicking on the 'Read Full Article' button, the user is directed to the web page from where the news was fetched from.
Farmers can directly contact the sellers and hence can save on money and can maximize profits. He/She has to enter details such as crop to be sold, asking price, quantity and contact details and these details can be forwarded to the sellers.
References:
● Crop yield prediction using machine learning algorithms - International Journal of Recent Technology and Engineering (IJRTE)
● Crop Condition Assessment using Machine Learning - International Journal of Recent Technology and Engineering (IJRTE)
● Open Government Data (OGD) Platform India
● Kaggle: Your Machine Learning and Data Science Community